Low Production - Shade vs Wiring vs Soiling Decision Tree

Why this matters

A customer reporting "my solar is producing less than it should" can be one of four things: actual production loss, customer misunderstanding the monitoring data, weather underperformance year over year, or a real hardware fault. Each gets different handling. Sending out a service truck to clean panels on a system that has a failed optimizer wastes the truck and leaves the customer believing the array works only because it was cleaned. The tree below sequences in the order field probability runs: easy data validation first, soiling and shading next, hardware last.

Symptom presentation

Pattern A: monitoring shows a percentage drop year-over-year on the same month, weather corrected. Pattern B: one or more panels showing significantly lower output in per-panel monitoring. Pattern C: production drops to zero or near-zero on a specific time of day repeatedly. Pattern D: production randomly drops then recovers, no pattern. Pattern E: customer perception of low production but data shows normal output.

Quick checks - 15 minute pass

1. Pull last 12 months production by month. Compare to same months previous year. A 3-5 percent decline is in the noise band of weather variability; 5-10 percent suggests soiling or new shading; above 10 percent suggests hardware.
2. Check the monitoring portal for active alarms or low-production flags.
3. Compare the customer's actual production to PVWatts modeled estimate at the same site, same azimuth, same tilt. A 15-25 percent underperformance vs PVWatts on day one is design issue; year-over-year decline is wear.
4. Walk the site at solar noon and again at customer-reported low-production time. Photograph any shading.
5. Read the inverter front panel or app for fault codes (Enphase Envoy, SolarEdge mySolarEdge, SMA SunSpec).

Isolation tree

Step 1 - Data validation (Pattern E)

Customer perception vs monitoring data:

• "My production was lower this month" - confirm month-over-month with weather correction. Cool cloudy months produce 30-50 percent less than peak; that is normal.
• "My system is undersized" - design issue, not service issue. PVWatts model shows what to expect.
• "My bill is higher" - utility may have raised rates, or customer added a new EV charger, hot tub, or other load. Compare consumption to historical, not just production.

If data validation shows normal production, the call closes here with education.

Step 2 - Per-panel data review (Pattern B)

If the system has per-panel monitoring (microinverters or optimizers):

• Sort panels by daily kWh ascending. Outliers below 80 percent of array average for the same day are real problems.
• Outlier on a south-facing panel without obvious shade: hardware. Failed microinverter or optimizer.
• Outliers in a clustered pattern (3 panels in a row): localized shade from a new obstruction (tree, addition).
• Outliers in a pattern that matches a known seasonal shadow path: shade.

Step 3 - String-level diagnosis (string inverter without optimizers)

String-level monitoring shows total output per string:

• One string at 50 percent of the other: blown fuse, failed module dragging the string, severe shade on one module.
• Both strings underproducing equally: inverter MPPT problem, soiling on the array, shading affecting both equally.
• Both strings showing erratic behavior: connector or wire issue.

Megger and IV-curve test by qualified tech if available; otherwise pull a single panel from the string, test independently, and reinsert to isolate.

Step 4 - Soiling diagnosis (covered in detail in companion article)

Soiling causes uniform reduction, slow seasonal build, recoverable with cleaning:

• Visual confirmation of dust, pollen, salt, bird streak.
• Production drop matches dry-season accumulation in arid climates.
• Production recovers within 24 hours of cleaning.

Step 5 - Shading diagnosis

New shading appears as a sharp drop at specific times of day, every day, on specific panels:

• Walk the site at the affected time. Photograph the shadow.
• Common sources: maturing tree (1-2 inch trunk diameter growth per year), new neighboring building, new HVAC condenser unit, new chimney cap, new satellite dish.
• Check for vent stack added in a recent reroof (a 2-inch plumbing stack throws a panel-wide shadow at low sun angles).

Shading must be solved or accepted. Trim if possible. Tigo TS4 add-on can mitigate string-level shading impact on legacy string systems.

Step 6 - Hardware fault diagnosis

Fault codes drive the path:

• Enphase microinverter fault on per-panel data: replace via warranty if under 25 years.
• SolarEdge optimizer offline or mismatched: replace optimizer.
• String inverter fault: check inverter manual for specific code. SMA "GROUND FAULT," Fronius "STATE 105," SolarEdge "Arc Detected" each indicate different issues.

Step 7 - DC wiring issues

Heating connectors are a known cause of long-term degradation:

• MC4 connectors at end of life (UV-degraded gaskets) heat up and present series resistance. IR thermography during peak production identifies hot connectors.
• DC arc faults (caught by AFCI in the inverter) point to bad connectors or chafed insulation.
• Conduit run under a hot roof develops insulation degradation; chafing at conduit entries causes shorts.

Remediation: replace any visibly degraded MC4 connectors with new field-installed connectors of matched manufacturer (do not mix Amphenol H4 with Stäubli MC4 - they look interchangeable, they are not).

Step 8 - Inverter age and end-of-life

String inverters have a 10-15 year life. A 12-year-old SMA Sunny Boy that has been working fine for a decade and starts showing production decline of 5-10 percent annually is at end of life. Capacitor aging in the inverter drops conversion efficiency. Plan replacement.

Microinverters typically have a 25-year warranty - failures within that window are replaced by manufacturer at parts cost. Customer may owe labor.

Confirming the diagnosis

PVWatts re-baseline production. Per-panel data sorted. Site photos of any shading. IR thermography on suspect connectors if available. Fault log from inverter.

Remediation

• Misperception, normal production: educate, document expected seasonal variation.
• Soiling: clean per separate decision tree.
• Shading: trim, mitigate with TS4 retrofit, or accept the loss.
• Hardware fault: replace failed component under warranty.
• DC wiring: replace degraded connectors and conduit entries.
• Inverter at end of life: plan replacement.

References

• IEC 61724-1 Photovoltaic System Performance Monitoring.
• NREL PVWatts Calculator and methodology.
• NEC 690.11 Arc-Fault Circuit Protection (residential PV).
• Enphase IQ Microinverter and Envoy User Guides.
• SolarEdge HD-Wave Inverter Installation and User Guide.